Oecologia (2003) 135:22–29
DOI 10.1007/s00442-002-1158-3
ECOPHYSIOLOGY
D. Glvez · R. W. Pearcy
Petiole twisting in the crowns of Psychotria limonensis:
implications for light interception and daily carbon gain
Received: 10 June 2002 / Accepted: 18 November 2002 / Published online: 7 February 2003
Springer-Verlag 2003
Abstract We used Y-plant, a computer-based model of
crown architecture, to examine the implications of leaf
reorientation resulting from petiole bending in Psychotria
limonensis (Rubiaceae) seedlings. During this reorientation process, bending of the petioles of lower leaves that
are potentially self-shaded by the upper leaves rotates the
lamina around the stem’s orthotropic axis so that selfshading is reduced. Simulations of daily light capture and
assimilation revealed a 66% increase in daily C gain due
to reorientation of the leaves as compared to simulations
where the leaves remained in their characteristic opposite
decussate pattern set by the phyllotaxy. This was due to
enhanced carbon (C) gain of the lower leaves because of
the reduction of shading by upper developing leaves in
the same vertical plane. The light signal for this
movement was experimentally examined by placing
leaf-shaped filters above already fully expanded leaves
and following the resulting shade-avoiding movements.
The filters were either neutral density shade cloth that
reduced the photon flux density (PFD) but did not alter
the red to far red ratio (R:FR) or a film that reduced the
PFD equivalently but also reduced the R:FR. Leaf
reorientation was much more rapid and complete under
the low R:FR as compared to the high R:FR indicating
involvement of a phytochrome photosensory system that
detected the presence of a shading leaf. Plants in gaps
were found to lack a reorientation response indicating that
the reorientation is specific to the shaded understory
environment.
Keywords Crown architecture · Petiole twisting ·
Self-shading · Light capture · Carbon dioxide assimilation
R. W. Pearcy ())
Section of Evolution and Ecology Division of Biological Sciences,
University of California,
1 Shields Avenue, Davis, CA, 95616, USA
e-mail: [email protected]
Tel.: +1-530-7521288
D. Glvez · R. W. Pearcy
Smithsonian Tropical Research Institute,
Apartado 2072, Balboa, Repfflblica de Panam
Introduction
The capture of light energy by a plant is governed by a
combination of leaf physiology and the display of those
leaves within the crown as set by the crown architecture.
For individual leaves, the orientation (azimuth and angle
of the surface) is a critical determinant of light capture
and leaf energy balance. At the plant crown level the
pattern of self-shading as determined by the spatial
arrangement of the leaves is an additional determinant of
light capture and photosynthesis. Diurnal solar tracking
and solar avoidance movements have been reported for a
large number of species in open habitats (Ehleringer and
Forseth 1980; Forseth and Ehleringer 1983; Kao and
Forseth 1992; Koller 1990). These studies have focused
on individual leaves with little consideration of the impact
of these reorientations on the performance of whole plant
crowns. There has also been little attention given to
understory plants. The spatial arrangement of leaves
within a crown depends on many factors including shoot
phyllotaxy (Niklas 1988; Takenaka 1994), branching
angles and frequencies (Fisher 1986; Waller and Steingraeber 1986), lengths of petioles and internodes (Takenaka 1994;Pearcy and Yang 1998; Yamada et al. 2000;
Takenaka et al. 2001), and leaf size and orientation. Most
research on crown architecture has focused on branching
patterns (Halle et al. 1978; Pickett and Kempf 1980;
Honda et al. 1997; Hatta et al. 1999), leaf display
(Chazdon 1985; Ackerly and Bazzaz 1995) or phyllotaxy
(Niklas 1988; Takenaka 1994) and their effects on light
capture. Only a few studies (Pearcy and Yang 1998;
Valladares and Pearcy 1998) have evaluated the effects of
architecture on photosynthetic carbon (C) gain or other
physiological attributes. While photosynthesis will be
related to light capture, especially in low light environments, the curvilinear response of photosynthesis to light
causes the relationship to depend on the interaction
between leaf physiological properties and light interception. Since the main function of leaf display is photosynthesis rather than light interception (Pearcy and
Valladares 1999; Valladares 1999) the consequences of
23
a particular architectural arrangement of leaves is best
evaluated by considering its impact on photosynthetic C
gain.
Monoaxial-stemmed plants present perhaps the simplest architectural leaf arrangement and consequently
have been more extensively investigated in terms of their
light capture and biomechanical properties than more
complex branched architectures (Givnish 1995; Niklas
1988; Takenaka 1994). Monoaxial plants with a few large
leaves are relatively common in tropical forest understories. Those few leaves must be able to maximize light
absorption for CO2 assimilation in order to maximize the
whole plant C gain necessary for growth in these
extremely shaded microenvironments (Pearcy 1999).
Simulations with model plants of defined architecture
have shown that phyllotaxy, leaf size, internode length
and petiole length interact in complex ways so that the
effect of any one character is not readily apparent without
considering it in the context of the other character states
(Niklas 1988, Takenaka 1994).
Psychotria limonensis. Krause (Rubiaceae), is an evergreen, broad-leaved understory shrub common on Barro
Colorado Island in Panama. It is locally abundant in
deeply shaded understory but also is found in treefall
gaps. Croat (1978) lists it as common in understories of
moist to wet forest from southern Mexico to northern
Colombia. Usually, it has a monoaxial crown, or rarely
branches. Stems on mature plants often lodge and root
adventitiously at nodes, which then can give rise to new
shoots. The resulting shoots are mostly orthotropic,
bearing large (200–300 mm long) broadly elliptic leaves
arranged in an opposite decussate phyllotaxy. The leaves
can persist for up to 4 years. Petioles are typically 40–
80 mm long. The opposite decussate leaf insertion pattern
potentially causes considerable self-shading in the vertical
projection since alternate leaf pairs overlap completely.
Since the small gaps inherent in otherwise closed
overstory canopies are much more frequent overhead
than to the side, this phyllotactic pattern is potentially
costly in terms of C gain. We observed, however, that
leaves on understory plants were often displaced from
their true phyllotactic angle, giving rise to a leaf display
superficially resembling that of a spiral rosette, which we
have termed a pseudo rosette. The leaf displacement,
which was due to petiole twisting, appeared to occur in
response to their shading by newly produced upper leaves.
This reorientation only occurred in understory plants;
those in gaps exhibited no displacement of their leaf
orientation from that created by an opposite decussate
phyllotaxy.
We used Y-plant, a three-dimensional computer based
model that reconstructs plant crown geometry and
simulates its light capture and C gain (Pearcy and Yang
1996), in order to characterize the dynamics of the leaf
reorientation and to determine its functional role in
whole-plant light capture and daily C gain. We also
examined the nature of the signals promoting reorientation with experimental manipulations of light quality and
quantity using specially constructed light filters mimicking leaves. We show that reorientation of lower leaves of
P. limonensis strongly enhances C gain in the understory
and that a phytochrome response to light quality [red to
far red ratio (R:FR)] is involved.
Materials and methods
Field site and plant material
All experiments were conducted in an area of secondary tropical
forest on Buena Vista Peninsula (BVP) that has been largely
undisturbed for at least 50 years in the Barro Colorado Natural
Monument, Panama (9º10'N, 79º51'W). BVP is located across the
Panama Canal (approximately 1 km) from the Smithsonian
Tropical Research Institute, Barro Colorado Island, Field Station
(BCI). BCI receives 2,650 mm precipitation per year and has a
strong, well-defined dry season typically lasting from December to
April.
Seedlings were grown from locally collected seed in 4-l pots in
a partially shaded growing house. The shade-cloth screened roof
resulted in a reduction in the daily photon flux density (PFD) to
approximately 12% of the total daily PFD in the open on BCI with
no change in the R:FR (R:FR=1.1). After germination, the plants
were thinned to a single plant per pot. After 2 months, 12 plants
were transferred to a fenced 77-m exclosure in the forest
understory and another 12 plants to a similar exclosure located in
an approximately 75-m2 tree fall gap that had been cleared of debris
and vegetation. Both exclosures were located nearby on BVP. The
plants at this time were approximately 20 cm tall and had three of
four leaf pairs. The older one or two leaf pairs on each plant were
removed so that only the upper two leaf pairs remained. Since these
two leaf pairs would not yet have exhibited any reorientation (i.e.,
the successive leaf pairs were at a 90 rotation to one another and
therefore not self-shaded), their reorientation could be followed
from the beginning as new leaves were produced above them. The
older leaves that were removed were small and shaded by the upper
leaves and, from simulations, make virtually no contribution to the
whole-plant C gain.
Simulation of light capture and C gain for understory plants
After removal of the lower leaves, those remaining on a plant were
carefully marked and the length, the azimuth and angle of the
normal to the surface, the azimuth of the midrib axis, and the
azimuth and angle of the petioles and internodes were recorded.
These initial data provided information on the orientation of leaves
prior to the development of new leaves above them. After 7 months,
two to three new leaf pairs had developed and the above
measurements were again made on both the original and the new
leaves. Hemispherical photographs were taken over each plant and
at a range of other understory sites with a Nikon model FM 35 mm
single lens reflex camera and Nikkor 8 mm fisheye lens. The photos
were analyzed with the program HemiView (version 2.1, Delta-T
Instruments, Cambridge) to determine the angular and azimuthal
distribution of gap fraction in the overstory canopy above each
plant, and the gap distribution along solar tracks for specific dates
as required for Y-plant. Leaf physiological parameters that are also
required to parameterize Y-plant were determined from light
response curves measured on other P. limonensis plants growing
nearby. The light response curves were constructed with a Li-Cor
6400 (Li-Cor, Lincoln, NE) photosynthesis system. Measurements
were made on the most recently fully expanded leaves of three
plants and values of light-saturated CO2 assimilation rates, the
apparent quantum yield and the dark respiration rate were
determined by least squares curve fitting to a rectangular hyperbola
light-response model (see Pearcy and Yang 1996). The mean values
from the three curve fits were then used as input into Y-plant. After
24
leaf expansion, the photosynthetic capacity of understory Psychotria spp. leaves in the BCI understory has been shown to decline
only slowly until senescence (K. Kitajima, personal communication).
After parameterization, Y-plant was used to reconstruct the 3dimensional architecture of each plant and then to calculate the
absorbed PFD and photosynthetic rate of each leaf, and by
summation over all leaves, the whole plant light absorption and
photosynthesis. Simulations in which the leaf laminas were
returned to their original position prior to reorientation were used
to study the effect of petiole reorientation on light capture and C
gain. This adjustment of the azimuths of the petioles and leaf
midribs input into Y-plant thus created the opposite decussate leaf
arrangement that would have occurred if it were not for petiole
reorientation. Comparisons of Y-plant simulations for the pseudo
rosette and the reconstructed opposite decussate architectures were
made under five different understory light regimes as simulated
from hemispherical photographs taken at different microsites in the
understory.
Simulation of light capture and C gain of gap-grown plants
In order to examine the potential consequences of petiole reorientation for plants in the gap light environments with higher PFD
levels, we manipulated both the leaf arrangement and the internode
length in Y-plant simulations. We measured the morphometric
parameters required for Y-plant on six opposite decussate 1-yearold P. limonensis plants growing in a gap garden. Leaf photosynthetic parameters were from measurements on similar plants in the
gap gardens following the same protocols as used for the understory
plants. We ran Y-plant simulations for opposite decussate or
pseudo-rosette leaf insertion patterns and for architectures in which
the internode lengths were varied from 15 to 200 mm. Y-plant
simulations utilized a hemispherical photo taken in the middle of
the garden (daily integral PFD=7.30 mol m-2 day-1).
Light quality effects on leaf reorientation
We utilized a shade cloth enclosure in the BCI growing house and
transferred into it twenty, 10-month old P. limonensis plants that
had initially been grown under the same conditions as those
transferred to the field exclosures. The shade cloth used reduced the
PFD at plant height in the enclosure to 80–130 mmol m-2 s-1 at
midday on clear days (approximately 5% of full sun, see Valladares
et al. 2000). At the time of transfer, the plants exhibited a typical
decussate leaf insertion pattern. After 14 days in the new light
environment, we randomly selected a fully expanded mature leaf
located below the terminal leaf pair on each plant, and above it we
placed a leaf-shaped wire frame supporting either a selective filter
(Energy Film; Gold Point ST7 STL-60, Panama) or a neutral
density filter (ND) constructed of shade cloth. The selective filter
will be identified as the LEAF filter since it mimicked the shading
by a leaf. It had a low transmission from 400 to 690 nm
wavelengths but cut on to a higher transmission in the near IR.
Thus, the transmitted PFD was reduced and the R:FR decreased to
values similar to those found in forest understories (Chazdon et al.
1996, Lee 1987). It had transmission characteristics in terms of
PFD and R:FR that were similar to those of a P. limonensis leaf
(Fig. 1). The ND filter, which had no effect on the R:FR, was
adjusted by altering the shade cloth combinations until one was
found that gave nearly the same reduction in PFD as the LEAF
filter. The PFD transmission and the R:FR under each filter were
determined using a quantum sensor (Li-189; Li-Cor) and a R:FR
sensor (SKR 110; Skye instruments, Powys, UK), respectively.
Following installation of the leaf filters, we recorded the midrib
azimuth of each leaf that was shaded by a filter with a compass
every 2 days for 27 days.
Fig. 1 Photon flux density (PFD) (black bars) and the red:far red
ratio (R:FR) (gray bars) light transmitted by the neutral density
(ND) filter, the energy film (LEAF) filter and a Psychotria
limonensis leaf. Each bar gives the mean and SE for five
measurements and bars with different letters are not significantly
different at the P=0.05 level
Results
After 7 months in forest understory, all plants had
naturally reoriented their lower leaves in order to avoid
the shadows cast by the newly expanded upper leaves.
This reorientation occurred in two different ways. In the
first, petioles of shaded leaves bent so as to rotate each
leaf blade around the axis in the same radial direction
(Fig. 2A). This created a new axis of symmetry offset 45
from those of the upper leaf pairs. Development of a new
leaf pair resulted in a similar rotation of a lower leaf pair
to occupy the remaining non- shaded space (Fig. 2B). At
this time, a rosette-like appearance (a pseudo rosette) of
leaf display was evident when the plants were viewed
from above. In the second mode of leaf reorientation,
which was less commonly observed, the two lower leaves
shaded by developing upper leaves rotated in opposite
radial directions around the axis (Fig. 2C). This initially
formed a non-symmetrical crown which could be later
returned to symmetry with an opposite rotation of other
leaf pairs.
Y-plant simulations with decussate versus pseudorosette architectures in the understory showed that the
latter had a significant advantage in terms of light capture
and C gain. Simulations were done for five different light
environments representing a range of daily PFDs characteristic of different microsites in the understory (Fig. 3)
using the leaf photosynthetic properties shown in Table 1.
Across all environments, daily C gain depended strongly
on the daily PFD ( r2=0.942 and 0.975 for the decussate
and pseudo-rosette plants, respectively) but plants with
the observed pseudo-rosette crowns had a significant C
gain advantage in comparison with simulated decussate
plants (ANOVA, P <0.005). Variations about the regression lines in Fig. 3 were due to the effects of the local
directionality of the light environments. The apparent
whole-plant light compensation point was much lower in
the pseudo-rosette as compared to the decussate plants
25
Fig. 3 Simulations of the daily CO2 assimilation in five different
understory light environments differing in daily PFD of the
naturally occurring pseudo-rosette crowns (m) and of the simulated
opposite decussate crowns (l) of P. limonensis. Each data point
gives the mean and SD for six plants
Fig. 2A–C Schematic diagrams illustrating the departures from
decussate phyllotaxis resulting from petiole twisting in P. limonensis. Leaf pairs are numbered in the sequence of development.
The original position of a leaf prior to rotation is shown by the
dashed outline. A Twisting of the petioles of lower leaves (1)
displaces the leaf blades outside the shadow of the newly expanded
leaves (3) forming a new axis of symmetry at 45 to the original
axis. B With the expansion of the next leaf pair (4), the leaves
below them (2) are displaced forming another axis of symmetry and
completing the change from a decussate to the pseudo-rosette
crown arrangement. C Less commonly, the shaded leaves (1) were
displaced in opposite directions after new leaves (3) expanded,
forming a non-symmetrical crown. Later rotations of leaf pairs in
the opposite direction could then return the crown to symmetry
(0.33 versus 0.66 mol m-2 day-1, respectively). In the
highest PFD microsite (1.2 mol m-2 day-1), the pseudorosette architecture exhibited a daily C gain nearly double
that of the decussate architecture. This difference was
clearly due to the increased C gain of the lower,
reoriented leaves (Fig. 4) as compared to those of the
simulated decussate leaf arrangement where the C gains
of the lower leaves were negative. As expected, upper
leaves had identical daily assimilation in both architectures.
P. limonensis plants in tree fall gaps did not exhibit the
reorientation characteristic of understory plants and instead
maintained a decussate leaf arrangement. These plants,
however, had much longer internodes than did the
understory plants, potentially allowing greater lateral light
penetration to lower leaves. We examined the interactions
between leaf arrangement and internode length in the gap
environments with Y-plant simulations to estimate daily
assimilation and water loss. Measurements of leaf photosynthesis revealed substantial acclimation to the higher
PFDs of the gap environment as compared to the
understory (Table 1). Therefore we used the photosynthetic
parameters derived from measurements on the gap plants
in these simulations. Since young seedlings produced much
shorter internodes (10–40 mm) than older plants (160–
210 mm), we varied internode length as well as leaf
arrangement (decussate versus pseudo-rosette). Simulations of daily assimilation revealed that pseudo-rosette
plants maintained a 20% C gain advantage (two factor
ANOVA, P<0.0001) over those with a decussate leaf
Table 1 Photosynthetic characteristicsa of Psychotria limonensis leaves from understory and gap light environments on Buena Vista
Peninsula
Amax
Quantum yield
-2 -1
Understory
Gap
a
-1
Curvature
Dark respiration
(mmol m s )
(mol CO2 mol photons)
(Dimensionless)
(mmol m-2 s-1)
3.28€0.11
10.44€0.59
0.063€0.006
0.068€0.004
0.769€0.041
0.559€0.204
0.206€0.032
0.471€0.151
The values shown are from a least squares fitting of the rectangular hyperbolic model:
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
fI ðfI þ Amax Þ2 4fqIAmax
AðI Þ ¼ Rd þ
2q
to photosynthetic light dependence responses where A(I) is the photosynthetic rate at a PFD of I, Rd is the dark respiration rate, Amax is the
light saturated photosynthetic rate, is the quanturm yield for CO2 uptake and is the curvature factor. Each is the mean and SE derived from
the fits to the light-response curves for three leaves
26
Table 2 Simulated absorbed photon flux density (PFD), assimilation and transpiration at midday under open sky conditions for
upper and lower leaves on P. limonensis plants with either opposite
decussate or pseudo-rosette architectures. Pseudo-rosette architectures were created from measured plants with decussate architectures by modifying the petiole and leaf orientations. Seedlings had
10-mm internode lengths and increasing leaf sizes from the base to
apex whereas the mature plants were simulated as having constant
leaf sizes and 200-mm internodes. Values for the upper leaves were
means for the two upper leaf pairs while values for the lower leaves
are for the six lowest leaf pairs
Decussate
Upper
Pseudo-rosette
Lower
-2
Fig. 4 Simulated daily CO2 assimilation for individual leaves
within the crown of pseudo-rosette plant (m) of P. limonensis, and
the reconstructed decussate plant (l) derived from it. Leaf position
is the node number from the lowest to the highest leaf pair
Fig. 5 Differences in daily CO2 assimilation of simulated plants
with decussate (l) and pseudo-rosette (m) leaf insertion patterns
on complete open gap light conditions. The horizontal dotted and
solid lines show the range of typical internode lengths during
seedling and mature plant stages, respectively. Each data point
represents the mean and SD for simulations for six plants
insertion pattern (Fig. 5) but internode length and the
interaction between internode length and phyllotaxy had no
significant effect (P=0.33 and P=0.06, respectively). Thus,
the advantage in daily assimilation was maintained over
the full range of internode lengths found in gap plants.
The lack of reorientation observed in gap plants is
present in spite of an apparent C gain advantage for
reorientation. Internode length interacts with phyllotactic
pattern and in open environments where there is significant lateral light, simulations have shown that longer
internodes can mitigate the impact of phyllotaxies that
result in self-shading by allowing more lateral light
interception (Niklas 1988; Takenaka 1994). However, in
the medium gaps for which this simulation was run, most
Upper
Lower
-1
Absorbed PFD at midday (mmol m s )
Seedling
1,775
163
1,844
Mature
1,825
675
1,811
Photosynthetic rate (mmol m-2 s-1)
Seedling
9.6
2.8
9.7
Mature
9.7
6.1
9.7
Transpiration rate (mmol m-2 s-1)
Seedling
0.060
0.024
0.061
Mature
0.061
0.053
0.061
1,170
1,020
7.5
7.4
0.045
0.051
of the PFD still originated near the zenith reducing any
impact of longer internodes. Examination of the simulated
daily assimilation of upper and lower leaves revealed the
expected pattern from the whole plant results, with the
lower self-shaded leaves on decussate plants having a
lower C gain than the comparable leaves on the pseudorosette plants and with the difference being largest for
seedlings (Table 2). This difference in assimilation was
accompanied by a 50% lower transpiration rate of the
lower leaves on seedlings.
We explored the role of light quality effects, specifically the R:FR, in the leaf reorientation by shading leaves
with either the ND or LEAF filters. Escape from the
shadows cast by the upper leaves should be maximized
when the leaf reaches a midpoint between the two
perpendicular symmetry axes of the outlines of the upper
leaves (see Fig. 2). This corresponds to a displacement
angle of 45, which we define as the maximum escape
angle. Indeed, shading with the LEAF filter, which
transmitted more FR, resulted in much more rapid and
complete movement towards this maximum escape angle
than observed with shading with a ND filter. A two-factor
repeated measures ANOVA revealed that the extent, the
rate of movement and the interaction term were all highly
significant (P<0.0001). During the first 6 days after
installation of the filters, leaf laminas under the LEAF
filter displaced 29, which is 65% of the maximum escape
angle (Fig. 6). In contrast, leaves under the ND filter
rotated only 6 or 13% of the maximum escape angle.
After 19 days reorientation was complete (Fig. 6), with
the leaves under the LEAF filter reaching their maximum
displacement from their true phyllotaxy (44€3), which
was not significantly different from the maximum escape
angle. In contrast, by 19 days leaves under ND leaf filters
had rotated only 34€4 and exhibited no further reorien-
27
Fig. 6 Displacement of leaf blades from the initial angle established by the shoot phyllotaxis following shading with either the
LEAF filter (m, low R:FR) or the ND filter (l, high R:FR). Each
data point gives the mean and SD for a leaf on each of five plants.
For abbreviations, see Fig. 1
tation. Thus, while leaves rotated under both types of
shading, the rate of rotation was much faster and more
complete under decreased R:FR. We observed that no
reorientation occurred in the absence of shading by either
a newly developing upper leaf or the artificial leaves. If a
developing leaf on an understory plant was removed then
the lower leaf that would have been shaded by it did not
reorient. However, the opposite lower leaf, which was
shaded by the newly developing opposite leaf, did
reorient as expected.
Discussion
The positioning of leaves around a shoot axis is a function
of phyllotaxy, the angle and length of the petioles, and any
reorientation of the lamina due to either diurnal movements or petiole or internode bending and twisting. The
latter are thought to be irreversible in the long term but are
clearly an important dynamic component that can maximize the efficiency of light capture. Previous studies of the
role of petioles in leaf display have focused on the effects
of petiole lengths and angles (Pearcy and Yang 1998;
Takenaka 1994; Takenaka et al.2001; Yamada et al. 2000)
and petiole deflection due to biomechanical load
(Niinemets 1998; Niklas 1992). To our knowledge, this
study is the first to quantify the effects of petiole twisting
for the C gain of plants and to identify signals involved in
the response. Previous studies of photomorphogenetic
responses (e.g., Marcuvitz and Turkington 2000) of
petioles have focused on reorientation and petiole extension during development of new leaves rather than
movements of already expanded leaves as occurs in P.
limonensis.
Niklas (1988) and Takanaka (1994) studied the central
role of phyllotaxy in efficient light capture and its
interaction with other morphological traits. Opposite leaf
arrangements have an advantage in that leaf area is
distributed on both sides of the node, minimizing the
support costs as compared to an equivalent leaf area
located on only one side. However, a disadvantage of the
opposite decussate architecture is that leaves on alternate
nodes are positioned directly above each other. Longer
internodes can reduce self-shading created by inefficient
phyllotaxies, especially when there is significant sidelight.
However, investment in additional internode length is
costly in terms of C that must be allocated to stems rather
than leaves. Since the understory vegetation is sparse in
forests where P. limonensis occurs, there is no particular
competitive advantage of elongating internodes as has been
demonstrated in denser herbaceous understories in temperate forests (Givnish 1982). Moreover, most of the PFD
available in understories originates from canopy gaps near
the zenith with little originating from the side, minimizing
the light capture advantages gained with longer internodes.
Alternate leaf arrangements in spiral phyllotaxies with
divergence angles approaching 137.5 are potentially more
efficient because they maximize the distance between
leaves that overlap in the vertical projection (Bell 1991;
Niklas 1988). Most plants with spiral phyllotaxies exhibit
divergence angles approaching this value, which maximizes light capture efficiencies. Reorientation of P.
limonensis leaves mimics a spiral architecture. That such
a reorientation is significant in terms of C gain is clearly
shown by the 66% increase in C gain at the normal range
of PFDs found in the understory habitat. This enhancement
of whole-plant C gain is likely to be crucial for survival of
P. limonensis plants in much darker microsites than would
be possible without reorientation.
Photomorphogenetic responses of plants are known to
be regulated by both blue and R:FR light photoreceptor
systems. Phototropic responses of stems have generally
been attributed to directional receipt of blue light via the
phototropin photoreceptor system (Casal 2000; Liscum
and Stowe-Evans 2000). However, R and FR light acting
through phytochrome have been shown to enhance the
response to blue light, causing phototropic curvature to be
more rapid (Liscum and Stowe-Evans 2000). Directional
light cues may be provided by blue light with FR
reception by phytochromes functioning to modulate the
magnitude of the curvature. It is also possible that
phytochrome alone can be a significant phototropic
receptor, especially in closed canopies, since significant
R:FR gradients can occur across tissues (Ballare 1994;
Ballare et al. 1991; Iino 1990). Due to the lack of suitable
filters we were unable explore the involvement of blue
light in leaf reorientation of P. limonensis. However, our
results showing greater and more rapid movements in
response to FR-enriched shading clearly support the
involvement of a phytochrome-mediated response as at
least part of the sensory transduction mechanism. Since
leaves are strong absorbers of R wavelengths but transmit
and reflect much more FR radiation, the R:FR serves as a
good indicator of shading by other leaves (Ballare 1999;
Smith 1995). Reorientation occurred in both neutral or
FR-enriched shade but was more rapid and greater in the
28
latter. This response is consistent with morphogenetic
responses to neutral and FR shade where for example
stem elongation occurs in response to both but is greater
in the latter (Lotscher and Nosberger 1997).
Although one of the classic responses to reduced R:FR
is stem elongation (Smith 1995), we saw no evidence of
stem elongation responses in P. limonensis in the shade.
Thus, the leaf reorientation response seems to be rather
specific and not a general simulation of phytochromemediated morphogenetic responses. Indeed, understory
plants have been shown to have much reduced stem
elongation responses to low R:FR as compared to highlight species (Morgan and Smith 1979), presumably
because stem elongation is a costly allocation pattern in
these light- and hence C-limited environments. R:FR is
already 0.2–0.3 in the understory giving a phytochrome
photoequilibrium strongly shifted to the Pr form. While
transmission through an additional leaf will lower it
further the scope for change is small. The reorientation
response in P. limonensis must therefore be quite
sensitive to small changes in R:FR.
Phytochrome is known to be localized both in leaves
and stems (Morgan et al. 1980). In our experiments only
the leaf lamina and not the petioles or stems were shaded
by the filters, suggesting that the leaf is the site of
perception of the R:FR signal involved in leaf reorientation. Thompson (1995) found that the petiole tip was the
site of photoreception for FR in red clover leaves. It is
likely that our filters also influence the light environment
of the petiole tip so we cannot rule out involvement of this
site as a photoreceptor. But clearly the R:FR received by
the leaf lamina was much more altered by the filter than
was the R:FR at the petiole tip. Moreover, phototropic
bending responses occur in response to light gradients
across the organ; it seems unlikely that a significant
lateral light gradient could be generated at the petiole tip.
Leaf development involves both lamina expansion and
petiole extension. The lamina of leaves that reoriented
were fully expanded at the time of development of the
upper shading leaf but it is possible that petioles may not
reach full extension until sometime later. Petiole elongation continued for a mean of 91 days in the tropical tree
species Macaranga gigantea even though lamina expansion required only 21 days (Yamada et al. 2000). This
example may be more extreme than the case for the much
shorter petioles of P. limonensis but it is possible that some
potential for extension remained at the time of reorientation. Except for diurnal leaf movements, which involve a
reversible swelling or shrinking of a pulvinus (Koller
1990), other phototropic responses occur in growing
organs. Blue light-mediated phototropic bending of stems
is caused by an inhibition of growth on the shaded
compared to the illuminated side, with little or no net effect
on overall growth, leading to a bending towards the light
source. If lamina are the site of perception of shading by an
upper developing leaf then transport of a signal to the
petioles is required. Greater shading on one leaf side than
the other could generate a differential signal. Translocation
of this signal via the pattern of venation could then carry it
differentially to one side of the petiole.
Although simulations of plants in medium gaps
showed that reorientation to the pseudo-rosette leaf
arrangement would yield a C gain advantage, no such
reorientation was actually observed. Plants in gaps
maintained the opposite decussate leaf arrangement set
by the phyllotaxy of P. limonensis. This lack of reorientation contrasts with findings of Ackerly and Bazzaz
(1995) for pioneer tree species seedlings, which oriented
their leaves in the direction of the prevailing diffuse light
in gaps. The reasons for no movement in P. limonensis are
not known but several possible factors could be involved.
First, in the high light of the gap environment, the signal
transduction mechanisms may have been saturated and
rendered inoperable. It is clear that in order for the
mechanisms to work in the shade environment they must
be very sensitive to very low photon fluences, which may
then limit their function in high light. The pioneer tree
species Ackerly and Bazzaz (1995) studied are adapted to
much higher light environments and the leaves appear to
orient during development rather than after full expansion. It may also be that a different signal transduction
mechanism, such as a blue light response acting alone, is
responsible for the leaf orientation during development.
Second, in this high light environment, C gain is unlikely
to be limiting, especially in a species principally adapted
for growth in understories. Therefore, a mechanism that
increases C gain may have little selective advantage.
Finally, simulations of transpiration of the lower leaves
show that it is higher in seedlings with a pseudo rosette as
compared to an opposite decussate leaf arrangement. This
may be an important cost of reorientation in gap
environments, especially in the long dry season characteristic of central Panama.
In conclusion, this study has shown that reorientation
of lower leaves of P. limonensis via petiole twisting
significantly enhances whole-plant C gain in deeply
shaded tropical forest understories. This reorientation
involves at least in part perception of shading by an upper
developing leaf through shifts in the R:FR of the light
received by the lower leaf. The lack of reorientation in
gap environments suggests that the response is specific to
shaded conditions where the advantage is greatest. Our
field observations indicate that leaf reorientation due to
petiole twisting is common in many species in both
shaded understories where it could function to enhance C
gain as we have shown with P. limonensis and in high
light environments where it may minimize receipt of
excess radiation, reducing stresses (Pearcy and Valladares
1999; Valladares and Pearcy 1998). Further studies are
warranted of both the mechanisms underlying these
responses and their consequences for light capture, C
gain and stress avoidance.
Acknowledgements This research was supported by NSF grant
IBN-9604424 and a Smithsonian Tropical Research Institute Short
Term Fellowship awarded to D. G. We thank Dr S. J. Wright and
the Smithsonian Tropical Research Institute for providing facilities
29
and support that made this research possible. Thanks to Drs Egbert
Leigh, Fernando Valladares, Timothy R. H. Pearson, Thomas
Kursar and Jim Dalling for their valuable comments and suggestions for the manuscript. D. G. thanks Prof. Claudia Peralta of
University of Panama for critical discussion during early stages of
the project and for comments on the manuscript.
References
Ackerly DD, Bazzaz FA (1995) Seedling crown orientation and
interception of diffuse radiation in tropical forest gaps. Ecology
76:1134–1146
Ballare CL (1994) Light gaps: sensing the light opportunities in
highly dynamic canopy environments. In: Caldwell MM,
Pearcy RW (eds) Exploitation of environmental heterogeneity
by plants: ecophysiological processes above-and belowground.
Academic Press, San Diego, Calif., pp 73–110
Ballare CL (1999) Keeping up with the neighbours: phytochrome
sensing and other signaling mechanisms. Trends Plant Sci
4:97–102
Ballare CL, Casal JJ, Kendrick RE (1991) Responses of lightgrown wild-type and long-hypocotyl mutant cucumber seedlings to natural and simulated shade light. Photochem Photobiol
54:819–826
Bell AD (1991) Plant form: an illustrated guide to flowering plant
morphology. Oxford University Press, Oxford
Casal JJ (2000) Phytochromes, cryptochromes, phototropin: photoreceptor interactions in plants. Photochem Photobiol 71:1–11
Chazdon RL (1985) Leaf display, canopy structure and light
interception of two understory palm species. Am J Bot
72:1493–1502
Chazdon RL, Pearcy RW, Lee DW, Fetcher N (1996) Photosynthetic responses of tropical forest plants to contrasting light
environments. In: Mulkey SS, Chazdon RL, Smith AP (eds)
Tropical forest plant ecophysiology. Chapman and Hall, New
York, pp 5–55
Croat TB (1978) Flora of Barro Colorado Island. Stanford
University Press, Stanford
Ehleringer J, Forseth I (1980) Solar tracking by plants. Science
210:1094–1098
Fisher JB (1986) Branching patterns and angles in trees. In: Givnish
TJ (ed) On the economy of plant form and function. Cambridge
University Press, Cambridge, pp 493–523
Forseth IN, Ehleringer JR (1983) Ecophysiology of two solar
tracking desert winter annuals. IV. Effect of leaf orientation on
calculated daily carbon gain and water use efficiency. Oecologia 58:10–18
Givnish TJ (1982) On the adaptive significance of leaf height in
forest herbs. Am Nat 120:353–381
Givnish TJ (1995) Plant stems: biomechanical adaptation for
energy capture and influence on species distribution. In:
Gartner BL (ed) Plant stems: physiology and functional
morphology. Academic Press, San Diego, Calif., pp 3–49
Halle F, Oldemann RAA, Tomlinson PB (1978) Tropical trees and
forests: an architectural analysis. Springer, Berlin Heidelberg
New York
Hatta H, Honda H, Fisher JB (1999) Branching principles
governing the architecture of Cornus kousa (Cornaceae). Ann
Bot 84:183–193
Honda H, Hatta H, Fisher JB (1997) Branch geometry in Cornus
kousa (Cornaceae): computer simulations. Am J Bot 84:745–
755
Iino M (1990) Phototropism — mechanisms and ecological
implications. Plant Cell Environ 13:633–650
Kao WY, Forseth IN (1992) Dirunal leaf movement, chlorophyll
fluorescence and carbon assimilation in soybean grown under
different nitrogen and water availabilities. Plant Cell Environ
15:703–710
Koller D (1990) Light-driven leaf movements. Plant Cell Environ
13:615–632
Lee DW (1987) The spectral distribution of radiation in two
neotropical rainforests. Biotropica 19:161–166
Liscum E, Stowe-Evans EL (2000) Phototropism: a “simple”
physiological response modulated by multiple interacting
photosensory-response pathways. Photochem Photobiol
72:273–282
Lotscher M, Nosberger J (1997) Branch and root formation in
Trifolium repens is influenced by the light environment of
unfolded leaves. Oecologia 111:499–504
Marcuvitz S, Turkington R (2000) Differential effects of light
quality, provided by different grass neighbours, on the growth
and morphology of Trifolium repens L. (white clover).
Oecologia 125:293–300
Morgan DC, Smith H (1979) A systematic relationship between
phytochrome-controlled development and species habitat, for
plants grown in simulated natural radiation. Planta 145:253–
258
Morgan DC, O’Brien T, Smith H (1980) Rapid photomodulation of
stem extension in light-grown Sinapsis alba L. Planta 150:95–
101
Niinemets U (1998) Adjustment of foliage structure and function to
a canopy light gradient in two co-existing deciduous trees.
Variability in leaf inclination angles in relation to petiole
morphology. Trees 12:446–451
Niklas KJ (1988) The role of phyllotactic pattern as a “developmental constraint” on the interception of light by leaf surfaces.
Evolution 42:1–16.
Niklas KJ (1992) Petiole mechanics, light interception by lamina,
and economy in design. Oecologia 90:518–526
Pearcy RW (1999) Responses of plants to heterogeneous light
environments. In: Pugnare FI, Valladares FV (eds) Handbook
of functional plant ecology. Decker, New York, pp 269–314
Pearcy RW, Valladares F (1999) Resource acquisition by plants:
the role of crown architecture. In: Press MC, Scholes J, Barker
M (eds) Plant physiological ecology. Blackwell Science,
Oxford, pp 45–66
Pearcy RW, Yang W (1996) A three-dimensional crown architecture model for assessment of light capture and carbon gain by
understory plants. Oecologia 108:1–12
Pearcy RW, Yang W (1998) The functional morphology of light
capture and carbon gain in the Redwood forest understorey
plant Adenocaulon bicolor Hook. Funct Ecol 12:543–552
Pickett STA, Kempf JS (1980) Branching patterns in forest shrubs
and understory trees in relation to habitat. New Phytol 86:219–
228
Smith H (1995) Physiological and ecological function within the
phytochrome family. Annu Rev Plant Physiol Plant Mol Biol
46:289–315
Takenaka A (1994) Effects of leaf blade narrowness and petiole
length on the light capture efficiency of a shoot. Ecol Res
9:109–114
Takenaka A, Takahashi K, Kohyama T (2001) Optimal leaf display
and biomass partitioning for efficient light capture in an
understorey palm, Licuala arbuscula. Funct Ecol 15:660–668
Thompson L (1995) Sites of photoperception in white clover. Grass
Forage Sci 50:259–262
Valladares F (1999) Architecture, ecology and evolution of plant
crowns. In: Pugnaire FI, Valladares F (eds) Handbook of
functional plant ecology. Decker, New York, pp 121–194
Valladares F, Pearcy RW (1998) The functional ecology of shoot
architecture in sun and shade plants of Heteromeles arbutifolia
M. Roem., a Californian chaparral shrub. Oecologia 114:1–10
Valladares F, Wright SJ, Lasso E, Kitajima K, Pearcy RW (2000)
Plastic phenotypic response to light of 16 congeneric shrubs
from a Panamanian rainforest. Ecology 81:1925–1936
Waller DM, Steingraeber DA (1986) Branching and modular
growth: theoretical models and empirical patterns. In: Cook R,
Buss L, Jackson JBC (eds) Biology of clonal organisms. Yale
University Press, New Haven, Conn., pp 35–70
Yamada T, Okuda T, Abdullah M, Awang M, Furukawa A (2000)
The leaf development process and its significance for reducing